A better understanding of how to form new -cells, and from where they might arise is of the utmost importance toward our goal of devising a strategy for -cell replacement therapy for diabetics. -cell replication appears to be the predominant mechanism underlying the generation of new -cells in young healthy animals, and perhaps in humans. However, neogenesis of -cells from non--cell sources may also play an important role, particularly in diabetics and in the elderly, where the capacity for -cell replicaion appears to be diminished. We feel that the pancreatic ducts have emerged as the most likely candidate for a non--cell source of new -cells. An anatomical association between the pancreatic ducts and the islets has been well-described for over 100 years, with most islets displaying some sort of contact with, or at least proximity to ducts. Here, however, in a genetically altered mouse model in which -cells do not proliferate after partial pancreatectomy, we describe a fairly dramatic sprouting of new ductal structures from existing large ducts after partial pancreatectomy. These sprouting ducts grow into and ramify within islets. Preliminary experiments strongly suggest that these intra-islet duct cells convert into islet cells. Interestinly, this ductal growth is largely suppressed if these same mice are bred into a background where proliferation of the islet cells after a pancreatectomy is restored. In addition, we found that suc islet-invading ducts are normally present transiently in young mice and in young humans, first appearing in mice at around two weeks of age, but then almost completely absent after eight weeks of age. There again (at least in mice) we found that those ductal cells in the young mice give rise to new insulin+ cells. In this proposal we will first strive to characterize the process y which these intra-islet duct structures arise, and in particular whether there is a specific subpopulation within the regular ductal network from which they arise. Second, we will study the molecular pathways that lead to their formation, and better define the phenotype of these invading ductal cells. Part of this analysis will entertain the possibility that these ductal structures arise from pancreatic ductal glands. Third, we will study the insulin+ cells that specifically form from these intra-islet ducts, and determine not only their precise phenotype, but also search for clues as to how they arose from the duct cells. We feel that a better understanding of these intra-islet ductal structures and the insulin+ cells they give rise to will have important implications for our ability to generate new -cells in the future, both in vitro an in vivo.